2 .\" Copyright (c) 2018 Stefan Sperling <stsp@openbsd.org>
4 .\" Permission to use, copy, modify, and distribute this software for any
5 .\" purpose with or without fee is hereby granted, provided that the above
6 .\" copyright notice and this permission notice appear in all copies.
8 .\" THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
9 .\" WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
10 .\" MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
11 .\" ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
12 .\" WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
13 .\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
14 .\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
21 .Nd Git repository format
23 A Git repository stores a series of versioned snapshots of a file hierarchy.
24 Conceptually, the repository's data model is a directed acyclic graph which
25 contains three types of objects as nodes:
28 The content of tracked files is stored in objects of type
33 object points to any number of such blobs, and also to other trees in
34 order to represent a hierarchy of files and directories.
38 object points to the root element of one tree, and thus records the
39 state of this entire tree as a snapshot.
40 Commit objects are chained together to form lines of version control history.
41 Most commits have just one successor commit, but commits may be succeeded by
42 an arbitrary number of subsequent commits so that diverging lines of version
43 control history, known as
46 A commit which precedes another commit is referred to as that other commit's
48 A commit with multiple parents unites disparate lines of history and is
53 Each object is identified by a SHA1 hash calculated over both the object's
54 header and the data stored in the object.
56 Loose objects are stored as individual files beneath the directory
58 spread across 256 sub-directories named after the 256 possible hexadecimal
59 values of the first byte of an object identifier.
60 The name of the loose object file corresponds to the remaining hexadecimal
61 byte values of the object's identifier.
63 A loose object file begins with a header which specifies the type of object
64 as an ASCII string, followed by an ASCII space character, followed by the
65 object data's size encoded as an ASCII number string.
66 The header is terminated by a
68 character, and the remainder of the file contains object data.
69 Loose objects files are compressed with
72 Multiple objects can be bundled in a
74 for better disk space efficiency and increased run-time performance.
75 The pack file format knows two additional types of objects in addition
76 to blobs, trees, and commits:
78 .It Offset Delta Objects
79 This object is represented as a delta against another object in the
81 This other object is referred to by its offset in the pack file.
82 .It Reference Delta Objects
83 This object is represented as a delta against another object in the
85 The other object is referred to by its SHA1 object identifier.
88 Pack files are self-contained and may not refer to loose objects or
89 objects stored in other pack files.
90 Deltified objects may refer to other deltified objects as their delta base,
91 forming chains of deltas.
92 The ultimate base of a delta chain must be an object of the same type as
93 the original object which is stored in deltified form.
95 Each pack file is accompanied by a corresponding
97 file, which lists the IDs and offsets of all objects contained in the
100 A reference associates a name with an object ID.
101 A prominent use of references is providing names to branches in the
102 repository by pointing at commit objects which represent the current
103 tip commit of a branch.
104 Because references may point to arbitrary object IDs their use
105 is not limited to branches.
107 The name is a UTF-8 string with the following disallowed characters:
115 [ (opening square bracket),
117 Additionally, the name may not contain the two-character sequences
120 Reference names may optionally have multiple components separated by
121 the / (slash) character, forming a hierarchy of reference namespaces.
124 reference namespace for internal use.
126 A symbolic reference associates a name with the name of another reference.
127 The most prominent example is the
129 reference which points at the name of the repository's default branch
132 References are stored either as a plain file within the repository,
137 file which contains one reference definition per line.
139 Any object which is not directly or indirectly reachable via a reference
140 is subject to deletion by Git's garbage collector.
142 .Bl -tag -width packed-refs -compact
144 A reference to the current head commit of the Git work tree.
145 In bare repositories, this files serves as a default reference.
147 Reference to original head commit.
148 Set by some Git operations.
150 Reference to a branch tip commit most recently fetched from another repository.
152 Legacy directory used by the deprecated Gogito Git interface.
154 Git configuration file. See
157 A human-readable description of the repository.
159 This directory contains hook scripts to run when certain events occur.
161 The file index used by
163 This file is not used by
169 Various configuration items.
171 Directory where reflogs are stored.
173 Loose and packed objects are stored in this directory.
175 A file which stores references.
176 Corresponding on-disk references take precedence over those stored here.
178 The default directory to store references in.
181 A typical Git repository exposes a work tree which allows the user to make
182 changes to versioned files and create new commits.
183 When a Git work tree is present, the actual repository data is stored in a
185 subfolder of the repository's root directory.
186 A Git repository without a work tree is known as a
190 does not make use of Git's work tree and treats every repository as if it
198 The Git repository format was initially designed by Linus Torvalds in 2005
199 and has since been extended by various people involved in the development
200 of the Git version control system.
202 The particular set of disallowed characters in reference names is a
203 consequence of design choices made for the command-line interface of
205 The same characters are disallowed by Got for compatibility purposes.
206 Got additionaly prevents users from creating reference names with
207 a leading - (dash) character, because this is rarely intended and
208 not considered useful.